Real-Time Risk, Is it Achievable?

Spending hours on Eurostar causes one to ponder on “stuff”. Real-time risk has been part of this “stuff”. Natixis has a case study that provides information on how it achieved real-time/intra-day risk coupled with P&L using various products, including Gigaspaces. Like most problems, the solution is either a complete custom build, a product or a hybrid. I suspect however that all real-time risk solutions need an in-memory cube to stand any chance of delivering the data in real-time. It’s a shame Natixis’s case study doesn’t provide a hint as to the latency from booking a trade to seeing the risk update on a traders desktop.

CEP also would appear to be important in the real-time risk space. Aleri Live OLAP is one such product that what offers real-time P&L. Aleri’s CEP engine is pretty good, but I don’t have a view on their Live OLAP server. I believe Aleri have had the Live OLAP product for a while, hence it would be curious to know if they have seen an uptick over the last year. Microsoft might also want to consider how it positions StreamInsight given that the group site inside the SQL Server team – StreamInsight coupled to Analysis Server would at least put them into the Aleri Live OLAP space.

Waters and Buy-Side Technology has an interesting view; “Although real-time risk management is still only possible in theory, the vendor has seen a large increase in interest in this area since the end of February. The firms looking to adopt this type of risk calculation are mostly large and mid-tier banks”.


~ by mdavey on November 13, 2009.

7 Responses to “Real-Time Risk, Is it Achievable?”

  1. The real question is “Real-Time Risk, Is it Desirable?”.

  2. Actually Natixis real time risk solution is built using QuartetFS’s ActivePivot realtime OLAP engine. We have recently published another case study with CMC markets where risk, PnLs and margin requirements are computed continuously ( ).

  3. I’ve doubts that the OLAP cube approach is viable when hard limit checks are required, i.e. don’t submit the order if you are going to break a limit. I don’t think OLAP implementations i’ve seen are not transactional, therefore limit breeches are possible, or have I completely misunderstood the technology.

    • Matt & Matt both make very good points about CEP & OLAP. In reality, hard pre-trade limits are best done in CEP where high speed, low latency checks can easily be done. ‘Real Time’ OLAP is best suited for post trade aggregation, monitoring and reporting where even if the CRO is breathing down your neck, providing the results within a second is still blindingly fast by their measures.

  4. There was a Java product called “JRisk” built by Application Networks that did this that I used on a project at a bank a few years back. It’s since been bought by Thompson Reuters and is now called “Top Office”. It had a real-time OLAP capability that was presented via the web-browser.
    We struggled with the product for a number of years until eventually the whole system was decommissioned. Over time the functionality improved but real time OLAP is certainly a very difficult thing to do well so I’m a little suspicious unless the solution has been around for a few years at a bank in production for a couple of years at least. JRisk suffered from problems like:
    – Server running out of memory due to too many users simultaneously openning too large cubes
    – Unreliable real-time updating
    – Difficult to define custom aggregations (e.g. adding up trade level P&L to the Book level when the trade level P&L’s are in different currencies and therefore need an FX conversion prior to adding)
    – Slow performance (up to 1 minute to recreate the view after the user changes the pivot at all)

    Does anyone know of a product that has matured beyond these kinds of growing pains?

  5. this is a very interesting article, i wonder what the future is hiding behaind its wings

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: